The Application of BP Neural Net Based on PCA in Software Risk Identification

摘要

According to large documents on the survey of software risk identification, we establish indicators system of software risk identification. In order to improve efficiency of sample data and reduce complexity of risk identification model, we use principal component analysis for noise reduction and, finally, take advantage of intelligent machine algorithm BP neural network establishing software risk identification model. An empirical analysis proves that the method has a higher value for reference and use.